OpenAI Launches o1-Pro – Its Most Powerful and Expensive AI Model Yet

Highlights

• OpenAI’s new o1-Pro model offers enhanced reasoning capabilities with more computational power for complex problems.
• It is priced at $150 per million tokens (input) and $600 (output).
• It includes a 200,000-token context window, image input support and structured outputs.
• The model is accessible only to select developers spending $5+ on OpenAI API services.
• It is unavailable for the Chat Completions API currently.

OpenAI has introduced its latest and most expensive AI model o1-Pro designed to offer superior reasoning capabilities. Available through the developer API, this model is a more powerful version of OpenAI’s o1, using significantly more computational power to deliver better responses to complex problems.

o1-pro now available in API @benhylak @literallyhimmmm @shl @joshRnold @samgoodwin89 @byamadaro1013 @adonis_singh @alecvxyz @StonkyOli @gabrielchua_ @UltraRareAF @yukimasakiyu @theemao @curious_vii

It uses more compute than o1 to provide consistently better responses. Available… pic.twitter.com/Iub6tp1NTi

— OpenAI Developers (@OpenAIDevs) March 19, 2025

“O1-pro in the API is a version of o1 that uses more computing to think harder and provide even better answers to the hardest problems. After getting many requests from our developer community, we’re excited to bring it to the API to offer even more reliable responses,” an OpenAI spokesperson told TechCrunch.

OpenAI o1-Pro – Key Features and Pricing

The o1-Pro model comes with a hefty price tag costing $150 per million tokens for input and $600 per million tokens for output. This makes it twice as expensive as GPT-4.5 and ten times the cost of the standard o1 model.

Despite this, OpenAI believes its enhanced performance will attract developers working on challenging tasks.

Some of its notable features include a 200,000-token context window, support for image inputs and structured output capabilities.

This makes o1-Pro ideal for applications requiring precise and reliable responses. However, early benchmarks show that while it performs slightly better than its predecessor in coding and math tasks, the improvements are incremental.

Open AI o1-Pro – Availability

For now, o1-Pro is only available to select developers who have spent at least $5 on OpenAI’s API services. It is primarily designed for AI agents with applications that can perform tasks autonomously through OpenAI’s new Responses API.

Meanwhile, developers using the Chat Completions API cannot access o1-Pro at this time.

FAQs

Q1. What is OpenAI’s o1-Pro model?

Answer. The o1-Pro is OpenAI’s most powerful and expensive AI model yet, designed for complex reasoning tasks with enhanced computational power.

Q2. What features and pricing does o1-Pro offer?

Answer. It includes a 200,000-token context window, image input support, structured outputs and costs $150 per million tokens for input and $600 for output.

Q3. Who can access o1-Pro and through which API?

Answer. It is available to select developers who have spent $5+ on OpenAI API services and is primarily designed for use with the Responses API.

Also Read: OpenAI Unveils GPT-4.5: ChatGPT’s Smarter, More Conversational Model Rolling Out Soon